US7000154B1 - System and method for fault detection and recovery - Google Patents

System and method for fault detection and recovery Download PDF

Info

Publication number
US7000154B1
US7000154B1 US10145449 US14544902A US7000154B1 US 7000154 B1 US7000154 B1 US 7000154B1 US 10145449 US10145449 US 10145449 US 14544902 A US14544902 A US 14544902A US 7000154 B1 US7000154 B1 US 7000154B1
Authority
US
Grant status
Grant
Patent type
Prior art keywords
system
object
event
hardware
software
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Expired - Fee Related, expires
Application number
US10145449
Inventor
Douglas E. LeDuc
John K. Lash
Nagendra V. Kolluru
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Intel Corp
Original Assignee
Intel Corp
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Grant date

Links

Images

Classifications

    • GPHYSICS
    • G06COMPUTING; CALCULATING; COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F11/00Error detection; Error correction; Monitoring
    • G06F11/07Responding to the occurrence of a fault, e.g. fault tolerance
    • G06F11/0703Error or fault processing not based on redundancy, i.e. by taking additional measures to deal with the error or fault not making use of redundancy in operation, in hardware, or in data representation
    • G06F11/0793Remedial or corrective actions
    • GPHYSICS
    • G06COMPUTING; CALCULATING; COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F11/00Error detection; Error correction; Monitoring
    • G06F11/07Responding to the occurrence of a fault, e.g. fault tolerance
    • G06F11/0703Error or fault processing not based on redundancy, i.e. by taking additional measures to deal with the error or fault not making use of redundancy in operation, in hardware, or in data representation
    • G06F11/0706Error or fault processing not based on redundancy, i.e. by taking additional measures to deal with the error or fault not making use of redundancy in operation, in hardware, or in data representation the processing taking place on a specific hardware platform or in a specific software environment
    • G06F11/0715Error or fault processing not based on redundancy, i.e. by taking additional measures to deal with the error or fault not making use of redundancy in operation, in hardware, or in data representation the processing taking place on a specific hardware platform or in a specific software environment in a system implementing multitasking
    • GPHYSICS
    • G06COMPUTING; CALCULATING; COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F11/00Error detection; Error correction; Monitoring
    • G06F11/07Responding to the occurrence of a fault, e.g. fault tolerance
    • G06F11/0703Error or fault processing not based on redundancy, i.e. by taking additional measures to deal with the error or fault not making use of redundancy in operation, in hardware, or in data representation
    • G06F11/0751Error or fault detection not based on redundancy
    • G06F11/0754Error or fault detection not based on redundancy by exceeding limits
    • G06F11/076Error or fault detection not based on redundancy by exceeding limits by exceeding a count or rate limit, e.g. word- or bit count limit

Abstract

An apparatus and method for automatically detecting and recovering from a fault in a microprocessor-based system. The apparatus and method utilizes a leaky bucket routine and an event handler procedure. The method may further use Object Oriented techniques that abstracts differences between hardware and software faults to allow for the development of a common framework.

Description

RELATED APPLICATION

This application claims the benefit of U.S. Provisional Patent Application Ser. No. 60/333,871 filed on Nov. 28, 2001 and which is incorporated herein by reference in its entirety.

BACKGROUND OF THE INVENTION

The present invention relates generally to fault detection and recovery and, more particularly, relates to a system and method for automatically detecting and recovering from software and/or hardware faults in microprocessor-based systems.

Microprocessor-based systems are used in an increasing number of applications, in part, because present day microprocessors are inexpensive and extremely powerful. Many of these systems are sophisticated and have complex software for driving the operation of the microprocessor and other hardware components. Since many of these systems, such as a router in a computer network, must operate continuously and unattended, the systems must be designed to operate in the presence of faults. These faults can be hardware faults or software faults resulting from hardware or software malfunctions.

In most microprocessor-based systems, fault detection and recovery is not implemented. In those rare cases where fault detection and recovery is implemented, the implementation is relatively primitive and informal. Specifically, it is typically left to the discretion of the hardware and software developers to design fault detection and recovery into their software processes which creates many problems. For example, any fault detection and recovery that does exist is tightly coupled and intertwined with the software process so re-use is difficult or impossible. This is especially true since software and hardware faults are typically handled by separate modules and not by one integrated module. Additional problems arise since many software processes are designed to exit when a fault occurs requiring the system to be manually restarted or rebooted to resume operation.

From the foregoing, it will be appreciated that a need exists for a more formal and comprehensive approach to hardware and fault detection and recovery. There is also a need for a fault detection and recovery method that can be easily re-used by any process or module in a system product. Finally, there is a need for fault recovery that is automatic in the sense that manual intervention is not required to recover from the fault.

SUMMARY OF THE INVENTION

In accordance with these needs, the present invention resides in an apparatus and method for automatically detecting and recovering from faults. To this end, faults may be reported as events. Processed events can then be used to increment an error count. Error counts may be based on an individual, family or system basis. The error counts may then be compared with thresholds and recovery functions can be performed if the thresholds are exceeded.

A better understanding of the objects, advantages, features, properties and relationships of the invention will be obtained from the following detailed description and accompanying drawings which set forth exemplary aspects of the invention and which are indicative of some of the ways in which the principles of the invention may be employed.

BRIEF DESCRIPTION OF THE DRAWINGS

FIG. 1 illustrates a block diagram of an exemplary microprocessor-based system that includes the fault detection and recovery subsystem of the present invention;

FIG. 2 illustrates a leaky bucket counter for use in the fault detection and recovery subsystem illustrated in FIG. 1;

FIG. 3 illustrates a flow chart diagram of an exemplary method for selecting a recovery function as a function of time for use in the fault detection and recovery subsystem illustrated in FIG. 1;

FIG. 4 illustrated an exemplary method of processing events for use in the fault detection and recovery system; and

FIGS. 5–8 illustrate exemplary system object hierarchies for use in fault detection and recovery.

DETAILED DESCRIPTION

Turning now to the figures, wherein like reference numerals refer to like elements, FIG. 1 illustrates a block diagram of an exemplary microprocessor-based system 100 that includes the fault detection and recovery subsystem 110 of the present invention. The microprocessor-based system 100 includes a hardware platform 120, an operating system 130, the fault detection and recovery subsystem 110 of the present invention, and software processes 140. The hardware platform includes a microprocessor. The operating system 130 can be any type of operating system and is embodied in software that resides on the hardware platform 120. The fault detection and recovery subsystem 110 is embodied in software that also resides on the hardware platform 120. The fault detection and recovery subsystem 110 can be divided into modules such as a first module 121 and a second module 122. The software processes 140 can be any type of software processes written in any programming language. The term “process” is used to indicate a series of software instructions that are recognized by the operating system 130 as a single unit. The software processes 140 are also resident on the hardware platform 120. Any number of software processes 140 can reside on the hardware platform 120, with the actual number limited by the operating system 130 and its resources. Four software processes 140 are shown in FIG. 1 for illustrative purposes only.

Turning to FIG. 2, there is illustrated a leaky bucket counter 200 for use in the fault detection and recovery subsystem illustrated in FIG. 1. In accordance with the leaky bucket process, which is preferably implemented in software, a counter 210 is incremented by the detection of a fault. The counter 210 also has a drain rate that decrements the counter 210. By decrementing the counter 210 at the drain rate, the number of detected faults will be smoothed over a period of time. Thus, if there are an unusual number of faults in a unit period of time, they will be detected. The leaky bucket counter 200 provides the ability to design a fault detection and recovery subsystem 110 that expects some faults to occur while providing for a way to perform recovery functions if the rate of faults detected is higher than expected. Adjusting the drain rate attribute allows for adjusting the number of faults in a unit of time that will be tolerated. In a microprocessor-based system 100 that requires zero fault tolerance, the drain rate can be set to zero.

Fault detection and recovery can be implemented partially as a library function called by the software process 141 and partially as processes within the fault detection and recovery subsystem 110 that are independent of the software processes 140 and thus can be used with any software processes 140. Accordingly, an application program interface (API) can be published for the fault detection and recovery subsystem 110 library functions. These library functions may be packaged as a runtime library and linked to the software process 141. In this manner, the software process 141 may call a fault detection and recovery subsystem 110 library function. The library function will then communicate with the independent processes of the fault detection and recovery subsystem 110.

Consideration for the length of time the microprocessor-based system has been running may also be utilized in the fault detection and recovery process. For example, more aggressive recovery functions may be desirable early on in the runtime of the microprocessor-based system 100, while less aggressive recovery functions may be utilized in later stages of the runtime of the microprocessor-based system 100. An example of this is shown in FIG. 3 that illustrates a flow chart diagram of an exemplary method for selecting a recovery function as a function of time. This method facilitates choosing a recovery function based on the time at which a fault occurs during the lifecycle of the microprocessor-based system 100.

In the example illustrated, the microprocessor-based system 100 is started up at step 305. The fault detection and recovery subsystem 110 is then initialized at step 310. At step 315, a first recovery function is selected that will be performed if a fault occurs. Any type of first recovery function may be utilized. For example, the first recovery function may be aggressive if a lot of things need to settle down within the microprocessor-based system 100 shortly after startup making the microprocessor-based system 100 fairly unstable during this time. Restarting, reinitializing, and reconfiguring the hardware and software within the microprocessor-based system 100 is an example of an aggressive recovery function.

At step 320, a timer is set to run for a first time interval. For example, fifteen minutes can be chosen for the first time interval. If a fault occurs within the software process 141 during the first time interval, the first recovery function will be performed by the fault detection and recovery subsystem 110. When the timer has expired at step 325, a second recovery function is selected at step 330 that will be performed if a fault occurs. Any type of second recovery function may be utilized. The second recovery function, for example, may be less aggressive than the first recovery function. For example, the second recovery function could comprise initializing the hardware and software in the microprocessor-based system 100.

If a predetermined number of time intervals have not yet elapsed at step 335, then the method 300 repeats starting at step 320 where a timer is set for a second time interval. If a fault occurs within the software process 141 during the second time interval, the second recovery function will be performed by the fault detection and recovery subsystem 110. The method 300 can be continued for as many time intervals or levels of recovery functions as are desired. For example, a third recovery function selected and then performed if a fault occurs in a third time interval could comprise initializing and reconfiguring only the software in the microprocessor-based system 100. A fourth recovery function selected and then performed if a fault occurs in a fourth time interval could comprise initializing the software in the microprocessor-based system 100. If the predetermined number of time intervals have elapsed at step 335, then a final recovery function is selected 340. If a fault occurs within the software process 141 during the remainder of the run time of the microprocessor-based system 100, the final recovery function will be performed by the fault detection and recovery subsystem 110.

Turning now to FIGS. 4–8, a system and method that utilizes event handling to detect and recover from faults is illustrated. As will be understood by those of skill in the art, an event is something that occurs to a software, hardware, or communication entity in a system. The event may be an indication that a condition (i.e., usually a problem or error) has occurred or is in progress or an indication that a condition has cleared (i.e., the condition is no longer present). Typical software events include interface errors, logic problems, system resource problems, real time problems, thrashing, supporting library failures, and sanity problems. Typical communication events include message time-outs, unexpected responses, receiving garbage/unrecognized data in a message, and underlying communication mechanism failures. Typical hardware events include hardware interrupts, hardware inconsistencies detected during polling, communication failures (typically detected through driver failures), and power/temperature/fan RPM anomalies. Any of these events may be detected through both hardware and software mechanisms. Still further, systems may address events encompassing zero or more of these areas. These events are handled through software that is dedicated to fault tolerance for that area.

It should be further appreciated that not all events are attributable to the element generating the event. In this regard, software failures account for the vast majority of all events encountered, even those reported by hardware elements. For example, drivers (software written to provide software with access to hardware in the system) may have flaws that cause the driver to fail to provide the access to the hardware it supports. On the other hand, the hardware itself may fail and report the same sort of failure event. Communication between software entities may experience problems due to problems in the sending software, problems in the communication mechanism (system calls, drivers, system hardware, or network), or problems in the receiving software. In this case, the problem may exist in any of the three areas. For this reason, it is desirable that events in all areas be handled through a common mechanism.

By way of example, FIG. 4 illustrates an overview of a process by which events may be handled and fault detection and recovery performed. In the illustrated example, sources of events may include, but are not limited to:

    • 1) Monitor processes—a process that monitors functions in system and reports events for abnormalities encountered. Examples of functions monitored are, but are not limited to, heartbeats from processes, memory usage of processes, or hardware.
    • 2) Normal software—as the software in the system processes it normal work, it reports events for errors encountered. Examples of errors are, but are not limited to, logic problems, communication problems, or data corruption.
    • 3) Driver/system call errors—drivers in the system (e.g. to access/control hardware, provide interfaces to other subsystems, etc.) and system calls (e.g. file input/output, sockets, allocate memory, etc.) may encounter errors for which events are reported. Examples of these errors are, but are not limited to, hardware not responding, file system full, or no memory available.
    • 4) Hardware interrupts—a hardware interrupt occurs when hardware failures occur. Examples of hardware failures are, but are not limited to, memory parity errors, power out of range, and temperature out of range.
    • 5) Other event detectors—Examples of the other event detectors would be a routine testing of various hardware, software or communication components in the system. This is done to periodically ensure the proper function of these components even when normal operation of the system does not require these functions. Therefore, pro-active actions can be taken to correct potential problems before those components are required.
      As illustrated, the handling of an event includes the steps of event analysis and event thresholding, which are described in greater detail hereinafter.

More specifically, for handing events, an approach is adopted wherein an object represents all entities of interest within the system. Objects are structured into a hierarchy. In addition, objects may be defined that represent entities external to that system with which the system is expected to establish communications. An exemplary system hierarchy is depicted in FIG. 5. Specifically, FIG. 5 illustrates a system that is decomposed into N hardware objects and a system software object (also be referred to as a family of software). The system software object is decomposed into the various M processes that run on that system. X objects are shown that represent entities external to the system with which the system may be expected to establish communications.

Within the system, a parent object is an object higher in the hierarchy than another related object (i.e., it can be reached from the object only following the solid connecting lines of FIG. 5 in the upward direction). Any object may have zero or more parent objects. The parent is decomposed into the objects that are below it. Therefore, in the example illustrated in FIG. 5, object “system” is the parent of objects “hardware object 1,” “hardware object N,” and “system software.” Object “system software” is a parent of objects “application process 1” and “application process M.” Note that object “hardware object 1” is not a parent of object “application process 1” even though it is higher in the hierarchy, since it cannot be reached from object “application process 1” only by following upward lines.

Similarly, a child object is an object lower in the hierarchy than another related object (i.e., it can be reached from the object only following the solid connecting lines of FIG. 5 in the downward direction). Any object may have zero or more children objects. The object is decomposed into the children objects below it, i.e., a child can be considered as “part of” its parent. Therefore, in the illustrated example, objects “hardware object 1,” “hardware object N,” and “system software” are all children of object “system.” Objects “application process 1” and “application process M” are both children of object “system software.” Note that object “application process 1” is not a child of “hardware object 1” even though it is lower in the hierarchy, since it cannot be reached from application process 1 only by following downward lines.

Still further, an associate object relationship is illustrated in FIG. 5 with the relationship being shown by a dashed line. Two objects are associated with each other if a special relationship exists between them that is not a parent-child relationship as described above. For example, an associate relationship exists when one object is expected to support another object. An associate relationship may cross any number of parent-child relationships. In the example illustrated in FIG. 5, an association exists between “application process 1” and “hardware object 1.” This relationship may be required since “application process 1” is not part of “hardware object 1” and “hardware object 1” can not be decomposed into “application process 1.” However, the system may require that “application process 1” provide all software interfaces to “hardware object 1.” If this were the case, an associate relationship exists between these two objects.

It is also possible to further decompose any of the objects shown in FIG. 5. For example, object “hardware object 1” might be decomposed into objects that are interfaces to the outside world (e.g., “Hardware object 1” is a board that provides access to two Ethernets and it contains two physical interfaces or connectors). Two objects may then be defined as children of “Hardware object 1,” each representing one of the Ethernet interfaces. FIG. 6 provides an example of this object model hierarchy.

For use in detecting and recovering from faults within the system, each object provides the ability to manage an event history. In turn, the event history list is managed by a leaky bucket list. As described previously, a leaky bucket is a counting device that has a specified drain rate. As events occur, the count in the leaky bucket is incremented and, as time elapses, the count in the leaky bucket is decremented as defined by its drain rate (e.g., 2 counts every minute). The drain rate is important, since events that occurred a week ago probably should not affect the count of events that occur today. A separate leaky bucket is preferably allocated for each type of event encountered for that object. Therefore, if a single object has experienced a communication time-out event and a software logic event, it would have two leaky buckets, one for each event type respectively. Once allocated, the leaky bucket is assigned an event type and a drain rate.

Importantly, to detect faults and provide fault tolerance and system stability, the objects and their family relationships are used to determine the most appropriate recovery action to take as a function of the events that are experienced by the objects of the system. To this end, an event handling (“EH”) module is defined that waits for events to occur. The EH module may be a single process or a collection of cooperating processes and preferably is the highest priority module/process(es) in the system. This enables the EH module to handle events without (or at least with minimal) delay. Under normal circumstances the EH module should be dormant and awakened when an event is reported.

For prioritizing the handling of events within the EH module, the events may be assigned a priority number based on the type of the event. If prioritization is attempted, priorities are preferably assigned based upon how much system instability is introduced by an event should processing of the event be delayed. The difference between system stability and ability to provide a major function of the system may also be taken into consideration. By way of example only, events can be prioritized as follows from highest to lowest: 1) power failure, fuse pulled/blown, board removed, operating system (OS) panic, or other events that imply the system will be completely unstable if processing of the event is delayed; 2) most internal hardware/interface, software, and communication events not included in priority 1 that provide some level of instability to the system; and 3) the remainder of events that include external hardware/interface events, and communication events to external entities (these are the least important events since the impact on system stability is minimal in face of communication problems to external entities). If multiple events occur simultaneously, the EH module should handle them in priority order and, within the same priority level, on a first-in-first-out (“FIFO”) order. By way of example, assuming that there are three event priorities as described, the EH module will process all priority 1 events in the order they occurred before priority 2 events, and all priority 2 events in the order they occurred before priority 3 events.

When the EH module is notified of an event, the EH module may handle the event by performing the following actions: 1) look for corruption in the event stream (e.g. sequence number violations) which requires corrective action on the event stream; 2) handle any throttling that may be necessary to control the rate of events into the EH module; 3) initialize data required by all EH module functions handling this event (e.g. time stamps, reporting object, etc.); 4) input event analysis to determine the list of objects that could have caused this event and how they could have caused this event; 5) event thresholding to analyze the objects on the list using event histories, thresholds, and other system state information to decide what action (if any) to take in response to this event; and 6) take the action chosen in the previous step. It is to be understood that the EH module need not perform all of these actions. Rather, certain actions may be omitted based upon the level of fault detection and correction desired.

To determine the list of objects that could have caused an event and how the event resulted from the object, the input event analysis step factors the event reported, the object the event is reported on, and the object hierarchies for the system. The developed list will be referred to hereinafter as the Object-Cause (“OC”) list. When developing the OC list, each suspect object placed in the OC list has an event cause that indicates how this suspect could have caused this event. The OC list should be sorted with the most likely suspect/cause at the top of the OC list, and the least likely at the bottom of the OC list. When a suspect object is identified, generally as a result of reporting an event or having an associate relationship with the reporting object pertaining to the performance of the task that generated the event, the parent object, grandparent object, and other ancestor objects of the suspect object (i.e., up the hierarchy) may also be placed on the OC list. This is because the suspect is “part of” the parent, which in turn is “part of” its parent, and so on. Therefore, each of the ancestors could have caused the problem. It should be noted that the higher in the hierarchy an object is the more general the cause that is assigned to the object. In this way, many events at a low level usually blame common ancestors with the same causes having the effect of accumulating cause counts for the common ancestors. By way of further example, the following paragraphs provide a variety of event examples to demonstrate how the OC list is constructed.

If two processes are communicating, and a timeout during a message exchange occurs, one of the processes reports the timeout event. Assume Process B sent the message to process A and Process B times out waiting for the reply from Process A. Table 1 shows the resulting OC list for a communication timeout event.

TABLE 1
Suspect Cause Reason
Process A Timeout Process took too long to process this message.
Process B insane Mismanaged its timer and a timeout fired
incorrectly
System Software The underlying communication mechanism in the
Software problem core system software/OS/communication library
may have lost the message sent or the response.
System Software Corruption has occurred in some set of the
corruption software on the system that is causing this
problem. The problem could be in the
communicating processes, in the system software,
in memory resources, shared libraries, etc.

In this example, the first column in Table 1 shows the suspect objects in the OC list, i.e., the object reporting the event, the associated object, and the ancestor objects. The second column provides a possible cause for the event, based upon the experiences of the system developers. The third column is a more detailed explanation of the how the suspect could have caused this event. It should be appreciated that the information that is reported within the third column is not part of the information required to proceed in this method—it is only provided here for supporting information. Furthermore, the suspect objects are ordered within the list, also based upon the experiences of the system developers, such that the first entry in the OC list is the most likely cause of this event. Note also that the ancestors of the first suspects appear in the list with very generalized causes.

A further example is provided with reference to FIG. 7 where a process is communicating with an external entity, receives an unrecognized message from the external entity, and reports the bad message as an event to the EH module. In this example, External entity X sends a message to Process A in the system and Process A determines that the message is corrupted. When the OC list is created, the association between Interface P and External entity X is considered since this association implies Hardware Q and its child Interface P exist in the system and are used for communicating to External entity X. Because of this association, the interface (and its parent) is added to list OC list which is illustrated in Table 2.

TABLE 2
Suspect Cause Reason
External Bad The external entity sent and invalid message or
Entity X message network connecting the system to the external
entity corrupted the message.
Process A insane Process has an internally corrupted the message
or process has moved into an error leg where it
is stuck a loop reporting events that are
not true.
System Software The underlying communication mechanism in
Software problem the core system software/OS/communication
library may have corrupted the message.
interface P Networking Something about the interface or it physical
corruption link with the network connecting it to External
entity X corrupts the message.
hardware Hardware Some hardware glitch, buffering problem, or
Q failure something with the processing of the message
corrupts the message.
System Software Corruption has occurred in some set of the
corruption software on the system that is causing this
problem. The problem could be in the
communicating processes, in the system
software, in memory resources, shared
libraries, etc.

As before, the first column in Table 2 shows the suspect objects in the OC list (i.e., the reporting process object, the object with which the reporting object was interacting when the event resulted, and the related and ancestral objects), the second column provides a cause for the event/problem, and the third column is an explanation of the how the suspect could have caused this event.

A still further example is provided for a logic error detected by a process that functions to indicate something is internally wrong with a software module. During the explanation of this example, reference may be had to the following sample of software which illustrates a potential logic error in the case of a lower bound check on a variable “a” failing:

Integer a;
If ( a > 5 )
Then
Do something
If ( a < 5 )
Then
Should never get here unless a logic error exists
End if
End if

If process A reaches the line “Should never get here unless a logic error exists,” an error event is reported to the EH module and the OC list in Table 3 may be created.

TABLE 3
Suspect Cause Reason
Process A Logic Process has corrupted its variable through some
error sort of logic error.
System Software The underlying communication mechanism may
Software problem have lost the message sent or the response.
System Software Corruption has occurred in some set of the
corruption software on the system that is causing this
problem. The problem could be in the
communicating processes, in the system software,
in memory resources, shared libraries, etc.

As with the previous example, the first column in Table 3 shows the suspect objects in the OC list (i.e., only the reporting object and its ancestral objects). The second column provides a possible cause for the problem. The optional third column is an explanation of the how the suspect could have caused this event.

Yet another example is provided, with reference to FIG. 8, in which hardware Q provides some function for the system and Process A is the interface (i.e., reads/writes) to hardware Q. This association is illustrated in FIG. 8 by the dashed line. In the event that Process A fails trying to read from Hardware Q and Process A reports the hardware failure even, the OC list in Table 4 may result.

TABLE 4
Suspect Cause Reason
hardware Hardware Something failed in hardware when trying to read
Q read it.
failure
System Hardware Problem in hardware in the system. Maybe clock
failure provided by the system skewed, power
glitched, or some interface device has failed.
Process A insane Process has an internally corrupted the message
or process has moved into an error leg where it is
stuck a loop reporting events that are not true.
System Software The underlying drivers in the core system
Software problem software/OS/communication library may have
failed causing the access failure.
System Software Corruption has occurred in some set of the
corruption software on the system that is causing this
problem. The problem could be in the
communicating processes, in the system software,
in memory resources, shared libraries, etc.

Once again, the first column in Table 4 shows the suspect objects in the OC list, the second column provides how the suspect could have caused this problem, and the optional third column includes an explanation of the how the suspect could have caused this event. Note that in this example, the object “system” is in the OC list twice—once for a hardware problem (as a parent to “Hardware Q” object) and once for a software problem (as a grandparent to “Process A” object).

Correlation of events leading to a root cause occurs naturally through the construction of the OC list and thresholding. As depicted through the examples found above, many events at a low level blame common ancestors in the object hierarchy with the same causes. When thresholding occurs, and the counts accumulate for the common ancestor, actions may not only be taken on a lower-level object, but may rather be taken on a higher-level object.

To decide what action, if any, to take in response to an event, the suspect objects of the OC list are analyzed using event histories, thresholds, and other system state information. To this end, a leaky bucket algorithm is used to hold (and drain at the specified rate) the count for a given object-cause combination. Furthermore, a predefined list of implementable actions is provided. By way of example only, and without limitation, possible actions for inclusion on the list are: 1) Do nothing; 2) Generate an alarm to the user of the system about the event; 3) Reset the hardware and re-initialize it and its driver (if one exists) to a clean state; 4) Switch hardware to a spare that is ready to take over in case the current hardware fails; 5) Reload a driver for a hardware object; 6) Kill process and restart it; 7) Kill process and start on a different processor; 8) Restart the system software; 9) Reboot the system; 10) escalate system recovery; and 11) Power down the system. Preferably, each predefined action is assigned a preference value that indicates which action would be more preferable to take than another. Typically the more system impacting the action the more preferable it is. Therefore, in the list provided above by way of example, the actions may be prioritized where “Do nothing” is the least preferred action and “Power down the system” is the most preferred action.

In the example provided, action 10 picks the current level of system recovery as defined in FIG. 3. The recovery would begin at reboot (action 9) during the first interval and de-escalate to restart system software (action 8) during the second interval. A distinction is drawn between action 10 and the actions 8 and 9—even though 10 may, at any point in time, be implemented by 8 or 9. This distinction exists, since some events require action 8 or 9 specifically to recover, whereas other events should take whatever the current action is for escalation based on the amount of time the system has been cycling and stable. The longer the system has been cycling, the less severe the action that is required to recover the system.

As will be appreciated, the aforementioned action list may vary from system to system depending on the mix of hardware, software, and external entities. For example, most systems will typically need at least the following actions if no additional hardware is needed/supported by the system: 1) Do nothing; 2) Generate an alarm to the user of the system about the event; 3) Kill process and restart it; 4) Restart the system software; 5) Reboot the system; 6) Escalate using a recovery procedure as illustrated by FIGS. 3; and 7) Power down the system. The escalation procedure for this example may begin at recovery (5) and as time goes on to recovery (4) and finally after a longer duration become recovery (3).

Similarly, for systems with some additional hardware, at least the following actions should be supported: 1) Do nothing; 2) Generate an alarm to the user of the system about the event; 3) Reset the hardware and re-initialize it and its driver (if one exists) to a clean state; 4) Reload a driver for a hardware object; 5) Kill process and restart it; 6) Restart the system software; 7) Reboot the system; 8) Escalate using the recovery procedure illustrated in FIGS. 3; and 9) Power down the system. The escalation procedure for this example may begin at recovery (7) and as time goes on to recovery (6) and finally after a longer duration become recovery (5). It should also be appreciated that the priority order of the actions that is set forth above may also be varied from system to system.

For responding to events, it is also preferred that predefined sets of thresholds be provided which are used to compare the current counts in the leaky buckets to determine an action to take from the list. The threshold may be defined as a series (1 or more) of entries. By way of example, the entry may contain the following information: 1) Count; 2) Action to be taken; 3) Set of system conditions that must be true before this threshold can be selected. (the set may be empty if no other system conditions are required to select this action). Table 5 is an example of a threshold entry set for a process having communication problems.

TABLE 5
Count Action Conditions
4 Kill and restart none
13 Kill and start on a Other processor exists that is ready to
different processor run this process

In the example provided in Table 5, the first column identifies the count that the leaky bucket must meet or exceed before the action is taken, the second column defines the action to take, and the third column defines any system conditions that must be true before this action can be selected. For example, assume the current count in the leaky bucket is 3. Since a value of 3 does not exceed any of the counts the default action is “Do nothing.” If the count in the leaky bucket is 6, the process will be killed and restarted. If the count is 14, but there is not another processor to schedule the process on, the only entry that the count exceeds and all conditions are true is the first entry—hence the action implemented would be “kill and restart” the process. If the count was 13 and there was another process available to schedule the process, the action would be “Kill and restart on a different processor.”

More specifically, for each entry in the OC list, the following steps are performed:

  • (1) If a leaky bucket to hold this combination of the object-cause does not exist:
    • (a) Allocate a leaky bucket to hold the counts for this object-cause combination; and
    • (b) Initialize the count and drain rates
  • (2) Increment the count in the leaky bucket for this combination object-cause. (The drain of the counts may be done now if appropriate, or may be managed independently of the method).
  • (3) Find the entry in the threshold list where both the current count in this leaky bucket has exceeded the threshold entry count and all conditions in the threshold entry are true.
    • (a) If no entry is found, the default action should be “do nothing.”
    • (b) If multiple entries are found the entry with the largest threshold count should be selected.
    • (c) If multiple threshold entries are still found, a selection criterion is applied (i.e. the first/last of the remaining matching entries is selected).
  • (4) Save the action of the selected threshold entry and the amount the leaky bucket count has exceeded the threshold entry count.

Once the action for each suspect in the OC list has been determined, one of the actions to be taken for the event is selected. This may be accomplished according to the following rules: 1) Select the action that is most preferable using the predefined preference values; 2) If rule “1” results in a multiple suspect/actions being selected, select the action for the suspect in the OC list where its count has exceeded its selected threshold entry count by the largest amount. (i.e., this suspect is selected because it has exceeded this threshold count for the larger number of events—and therefore has needed this action for a longer period of time); and 3) In the event after both rule “1” and rule “2” have been applied and multiple suspects with the same action still exist, the order they appear in the OC list should be used to resolve the conflict. (i.e., the suspects were added into the OC list in the order that they most likely caused the event).

While specific aspects of the present invention have been described in detail, it will be appreciated by those skilled in the art that various modifications and alternatives to those details could be developed in light of the overall teachings of the disclosure. Accordingly, the particular arrangement disclosed is meant to be illustrative only and not limiting as to the scope of the invention which is to be given the fill breadth of the appended claims and any equivalents thereof.

Claims (20)

1. A method for automatically detecting and recovering from a fault in a microprocessor-based system, comprising:
reporting the fault as an event;
processing the event including thresholding the event and co-relating the event to a cause;
determining a recovery action as a function of the thresholding, the co-relating, and an elapsed time the system has been running, the recovery action being used to perform one or more of a restart of the system, cleanup of memory and data prior to restart, hardware resets to hardware modules or sub-assemblies, or releasing of resources that are marked unavailable due to the faulty behavior, wherein the recovery action is more aggressive in initial stages and less aggressive in later stages; and
performing the recovery action.
2. A method for detecting and recovering from faults in a system represented by an object hierarchy, comprising:
in response to an event indicative of a fault condition, creating a list of object-cause pairs relevant to the event;
incrementing a count in a leaky bucket associated with each object-cause pair; and
if the count of the leaky bucket exceeds a plurality of thresholds, performing an action that has been associated with a highest threshold.
3. The method as recited in claim 2, wherein creating the list comprises identifying a first object from which the event originated and including on the list the first object and ancestral objects of the first object.
4. The method as recited in claim 3, wherein creating the list comprises identifying a second object having an associate relationship with the first object in connection with a task that relates to the event and including on the list the second object and ancestral objects of the second object.
5. The method as recited in claim 4, further comprises ordering the object-cause list based on a predetermined priority.
6. The method as recited in claim 2, wherein the plurality of thresholds are defined for at least one leaky bucket, each threshold having an associated action to be performed.
7. The method as recited in 6, wherein the actions are selected from a list consisting of do nothing; generate an alarm to the user of the system about the event; reset the hardware and re-initialize it and its driver to a clean state; switch hardware to a spare that is ready to take over in case the current hardware fails; reload a driver for a hardware object; kill process and restart it; kill process and start on a different processor; restart the system software; reboot the system; and power down the system.
8. The method as recited in claim 2, further comprising determining if a condition within the system is true before performing an action if the threshold is exceeded.
9. A computer-readable media having instructions for detecting and recovering from faults in a system represented by an object hierarchy, the instructions performing steps comprising:
in response to an event indicative of a fault condition, creating a list of object-cause pairs relevant to the event;
incrementing a count in a leaky bucket associated with each object-cause pair; and
if the count of the leaky bucket exceeds a plurality of thresholds, performing an action that has been associated with a highest threshold.
10. The media as recited in claim 9, wherein creating the list comprises identifying a first object from which the event originated and including on the list the first object and ancestral objects of the first object.
11. The media as recited in claim 10, wherein creating the list comprises identifying a second object having an associate relationship with the first object in connection with a task that relates to the event and including on the list the second object and ancestral objects of the second object.
12. The media as recited in claim 11, wherein the instruction further order the object-cause list based on a predetermined priority.
13. The media as recited in claim 9, wherein the plurality of thresholds are defined for a least one leaky bucket each threshold having an associated action to be performed.
14. The media as recited in 13, wherein the actions are selected from a list consisting of do nothing; generate an alarm to the user of the system about the event; reset the hardware and re-initialize it and its driver to a clean state; switch hardware to a spare that is ready to take over in case the current hardware fails; reload a driver for a hardware object; kill process and restart it; kill process and start on a different processor; restart the system software; reboot the system; and power down the system.
15. The media as recited in claim 9, wherein the instructions further determine if a condition within the system is true before performing an action if the threshold is exceeded.
16. A method for automatically detecting and recovering from a fault in a microprocessor-based system, comprising:
reporting the fault as an event;
processing the event including thresholding the event and co-relating the event to a cause;
determining a recovery action as a function of the thresholding, the co-relating, and an elapsed time the system has been running, wherein recovery is more aggressive in initial stages and less aggressive in later stages, the recovery actions being used to perform one or more of a restart of the system, cleanup of memory and data prior to restart, hardware resets to hardware modules or sub-assemblies, or releasing of resources that are marked unavailable due to the faulty behavior; and
performing the recovery action.
17. A method for detecting and recovering from faults in a system represented by an object hierarchy, comprising:
in response to an event indicative of a fault condition, creating a list of object-cause pairs relevant to the event;
incrementing a count in a leaky bucket associated with each object-cause pair; and
if the count of the leaky bucket exceeds a plurality of thresholds, performing an action that has been associated with a highest threshold, wherein the plurality of thresholds are defined for at least one leaky bucket, each threshold having an associated action to be performed.
18. A method for detecting and recovering from faults in a system represented by an object hierarchy, comprising:
in response to an event indicative of a fault condition, creating a list of object-cause pairs relevant to the event;
incrementing a count in a leaky bucket associated with each object-cause pair; and
if the count of the leaky bucket exceeds a plurality of thresholds, performing an action elected from a list consisting of do nothing; generate an alarm to the user of the system about the event; reset the hardware and re-initialize it and its driver to a clean state; switch hardware to a spare that is ready to take over in case the current hardware fails; reload a driver for a hardware object; kill process and restart it; kill process and start on a different processor; restart the system software; reboot the system; and power down the system, wherein the plurality of thresholds are defined for at least one leaky bucket, each threshold having an associated action to be performed.
19. A computer-readable media having instructions for detecting and recovering from faults in a system represented by an object hierarchy, the instructions performing steps comprising:
in response to an event indicative of a fault condition, creating a list of object-cause pairs relevant to the event;
incrementing a count in a leaky bucket associated with each object-cause pair; and
if the count of the leaky bucket exceeds a plurality of thresholds, performing an action that has been associated with a highest threshold, wherein the plurality of thresholds are defined for a least one leaky bucket each threshold having an associated action to be performed.
20. A computer-readable media having instructions for detecting and recovering from faults in a system represented by an object hierarchy, the instructions performing steps comprising:
in response to an event indicative of a fault condition, creating a list of object-cause pairs relevant to the event;
incrementing a count in a leaky bucket associated with each object-cause pair; and
if the count of the leaky bucket exceeds a plurality of thresholds, performing an action selected from a list consisting of do nothing; generate an alarm to the user of the system about the event; reset the hardware and re-initialize it and its driver to a clean state; switch hardware to a spare that is ready to take over in case the current hardware fails; reload a driver for a hardware object; kill process and restart it; kill process and start on a different processor; restart the system software; reboot the system; and power down the system, wherein the plurality of thresholds are defined for a least one leaky bucket each threshold having an associated action to be performed.
US10145449 2001-11-28 2002-05-13 System and method for fault detection and recovery Expired - Fee Related US7000154B1 (en)

Priority Applications (2)

Application Number Priority Date Filing Date Title
US33387101 true 2001-11-28 2001-11-28
US10145449 US7000154B1 (en) 2001-11-28 2002-05-13 System and method for fault detection and recovery

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
US10145449 US7000154B1 (en) 2001-11-28 2002-05-13 System and method for fault detection and recovery
US11290777 US7386764B2 (en) 2001-11-28 2005-11-29 System and method for fault detection and recovery

Publications (1)

Publication Number Publication Date
US7000154B1 true US7000154B1 (en) 2006-02-14

Family

ID=35768080

Family Applications (2)

Application Number Title Priority Date Filing Date
US10145449 Expired - Fee Related US7000154B1 (en) 2001-11-28 2002-05-13 System and method for fault detection and recovery
US11290777 Active 2022-08-04 US7386764B2 (en) 2001-11-28 2005-11-29 System and method for fault detection and recovery

Family Applications After (1)

Application Number Title Priority Date Filing Date
US11290777 Active 2022-08-04 US7386764B2 (en) 2001-11-28 2005-11-29 System and method for fault detection and recovery

Country Status (1)

Country Link
US (2) US7000154B1 (en)

Cited By (26)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20050195840A1 (en) * 2004-03-02 2005-09-08 Steven Krapp Method and system for preventing denial of service attacks in a network
US20060048019A1 (en) * 2004-08-26 2006-03-02 Canon Kabushiki Kaisha Image input/output device, information processing method, storage medium stored with program readable by computer, and program
US20060230306A1 (en) * 2005-03-31 2006-10-12 Richards Paul M Apparatus, system, and method for facilitating monitoring and responding to error events
US20060248531A1 (en) * 2005-03-31 2006-11-02 Oki Electric Industry Co., Ltd. Information processing device, information processing method and computer-readable medium having information processing program
US20070055915A1 (en) * 2005-09-07 2007-03-08 Kobylinski Krzysztof R Failure recognition, notification, and prevention for learning and self-healing capabilities in a monitored system
US20070055914A1 (en) * 2005-09-07 2007-03-08 Intel Corporation Method and apparatus for managing software errors in a computer system
US20070177523A1 (en) * 2006-01-31 2007-08-02 Intec Netcore, Inc. System and method for network monitoring
US20070240207A1 (en) * 2004-04-20 2007-10-11 Ecole Polytechnique Federale De Lausanne (Epfl) Method of Detecting Anomalous Behaviour in a Computer Network
US20070245313A1 (en) * 2006-04-14 2007-10-18 Microsoft Corporation Failure tagging
US20080184079A1 (en) * 2007-01-31 2008-07-31 Microsoft Corporation Tracking down elusive intermittent failures
US20080184075A1 (en) * 2007-01-31 2008-07-31 Microsoft Corporation Break and optional hold on failure
US20080256149A1 (en) * 2007-04-13 2008-10-16 Nikhil Bansal System and method for dependent failure-aware allocation of distributed data-processing systems
US20090300770A1 (en) * 2002-09-18 2009-12-03 Rowney Kevin T Mechanism to search information content for preselected data
US20100152894A1 (en) * 2008-12-15 2010-06-17 Samsung Electronics Co., Ltd. Robot and recovery method thereof
US20100162347A1 (en) * 2008-12-22 2010-06-24 Ian Barile Adaptive data loss prevention policies
US20110099638A1 (en) * 2002-09-18 2011-04-28 Chris Jones Method and apparatus to report policy violations in messages
US20110202797A1 (en) * 2010-02-12 2011-08-18 Evgeny Mezhibovsky Method and system for resetting a subsystem of a communication device
EP2642390A1 (en) * 2012-03-20 2013-09-25 BlackBerry Limited Fault recovery
US8566305B2 (en) 2002-09-18 2013-10-22 Symantec Corporation Method and apparatus to define the scope of a search for information from a tabular data source
US8751506B2 (en) 2003-05-06 2014-06-10 Symantec Corporation Personal computing device-based mechanism to detect preselected data
US8813176B2 (en) 2002-09-18 2014-08-19 Symantec Corporation Method and apparatus for creating an information security policy based on a pre-configured template
US8826443B1 (en) 2008-09-18 2014-09-02 Symantec Corporation Selective removal of protected content from web requests sent to an interactive website
US8935752B1 (en) 2009-03-23 2015-01-13 Symantec Corporation System and method for identity consolidation
US9026842B2 (en) 2012-03-20 2015-05-05 Blackberry Limited Selective fault recovery of subsystems
US9235629B1 (en) 2008-03-28 2016-01-12 Symantec Corporation Method and apparatus for automatically correlating related incidents of policy violations
US9515998B2 (en) 2002-09-18 2016-12-06 Symantec Corporation Secure and scalable detection of preselected data embedded in electronically transmitted messages

Families Citing this family (21)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2004038700A3 (en) 2002-10-23 2004-08-12 Onaro Method and system for validating logical end-to-end access paths in storage area networks
US7961594B2 (en) * 2002-10-23 2011-06-14 Onaro, Inc. Methods and systems for history analysis for access paths in networks
US7266726B1 (en) * 2003-11-24 2007-09-04 Time Warner Cable Inc. Methods and apparatus for event logging in an information network
US8302111B2 (en) 2003-11-24 2012-10-30 Time Warner Cable Inc. Methods and apparatus for hardware registration in a network device
US8078669B2 (en) 2004-02-18 2011-12-13 Time Warner Cable Inc. Media extension apparatus and methods for use in an information network
EP1943594A4 (en) * 2005-09-27 2009-12-16 Onaro Methods and systems for validating accessibility and currency of replicated data
US7757030B2 (en) * 2005-12-16 2010-07-13 Microsoft Corporation Simulating hardware dynamic partitioning capabilities
US8370818B2 (en) 2006-12-02 2013-02-05 Time Warner Cable Inc. Methods and apparatus for analyzing software interface usage
US7535349B2 (en) * 2006-12-08 2009-05-19 International Business Machines Corporation Determining root cause for alarm in processing system
US8826032B1 (en) 2006-12-27 2014-09-02 Netapp, Inc. Systems and methods for network change discovery and host name resolution in storage network environments
US8332860B1 (en) 2006-12-30 2012-12-11 Netapp, Inc. Systems and methods for path-based tier-aware dynamic capacity management in storage network environments
US9042263B1 (en) 2007-04-06 2015-05-26 Netapp, Inc. Systems and methods for comparative load analysis in storage networks
EP2166452B1 (en) * 2007-06-04 2012-08-01 Fujitsu Limited Restart support program and restart support device
US20090103916A1 (en) * 2007-10-18 2009-04-23 Csc Holdings, Inc. Optical chassis monitoring
US8099637B2 (en) * 2007-10-30 2012-01-17 Hewlett-Packard Development Company, L.P. Software fault detection using progress tracker
US8392763B2 (en) * 2009-11-24 2013-03-05 International Business Machines Corporation Technique for estimation of confidence interval for probability of defect rediscovery
JP5517301B2 (en) * 2010-08-23 2014-06-11 ルネサスエレクトロニクス株式会社 Data processing system
US20120221884A1 (en) * 2011-02-28 2012-08-30 Carter Nicholas P Error management across hardware and software layers
US8516506B2 (en) * 2011-03-29 2013-08-20 Denso International America, Inc. Method and system for restoring an application in a dynamically linked environment
CN105577315A (en) * 2014-10-08 2016-05-11 深圳市中兴微电子技术有限公司 Link state control method and device
US9612894B2 (en) 2015-06-01 2017-04-04 Nxp Usa, Inc. Detector for high frequency interrupts

Citations (10)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US5528516A (en) * 1994-05-25 1996-06-18 System Management Arts, Inc. Apparatus and method for event correlation and problem reporting
US5864662A (en) * 1996-06-28 1999-01-26 Mci Communication Corporation System and method for reported root cause analysis
US6006016A (en) * 1994-11-10 1999-12-21 Bay Networks, Inc. Network fault correlation
US6327677B1 (en) * 1998-04-27 2001-12-04 Proactive Networks Method and apparatus for monitoring a network environment
US6414595B1 (en) * 2000-06-16 2002-07-02 Ciena Corporation Method and system for processing alarm objects in a communications network
US6421719B1 (en) * 1995-05-25 2002-07-16 Aprisma Management Technologies, Inc. Method and apparatus for reactive and deliberative configuration management
US6560725B1 (en) * 1999-06-18 2003-05-06 Madrone Solutions, Inc. Method for apparatus for tracking errors in a memory system
US6604208B1 (en) * 2000-04-07 2003-08-05 Telefonaktiebolaget Lm Ericsson (Publ) Incremental alarm correlation method and apparatus
US6738933B2 (en) * 2001-05-09 2004-05-18 Mercury Interactive Corporation Root cause analysis of server system performance degradations
US6754854B2 (en) * 2001-06-04 2004-06-22 Motorola, Inc. System and method for event monitoring and error detection

Patent Citations (10)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US5528516A (en) * 1994-05-25 1996-06-18 System Management Arts, Inc. Apparatus and method for event correlation and problem reporting
US6006016A (en) * 1994-11-10 1999-12-21 Bay Networks, Inc. Network fault correlation
US6421719B1 (en) * 1995-05-25 2002-07-16 Aprisma Management Technologies, Inc. Method and apparatus for reactive and deliberative configuration management
US5864662A (en) * 1996-06-28 1999-01-26 Mci Communication Corporation System and method for reported root cause analysis
US6327677B1 (en) * 1998-04-27 2001-12-04 Proactive Networks Method and apparatus for monitoring a network environment
US6560725B1 (en) * 1999-06-18 2003-05-06 Madrone Solutions, Inc. Method for apparatus for tracking errors in a memory system
US6604208B1 (en) * 2000-04-07 2003-08-05 Telefonaktiebolaget Lm Ericsson (Publ) Incremental alarm correlation method and apparatus
US6414595B1 (en) * 2000-06-16 2002-07-02 Ciena Corporation Method and system for processing alarm objects in a communications network
US6738933B2 (en) * 2001-05-09 2004-05-18 Mercury Interactive Corporation Root cause analysis of server system performance degradations
US6754854B2 (en) * 2001-06-04 2004-06-22 Motorola, Inc. System and method for event monitoring and error detection

Cited By (41)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20090300770A1 (en) * 2002-09-18 2009-12-03 Rowney Kevin T Mechanism to search information content for preselected data
US20110099638A1 (en) * 2002-09-18 2011-04-28 Chris Jones Method and apparatus to report policy violations in messages
US8566305B2 (en) 2002-09-18 2013-10-22 Symantec Corporation Method and apparatus to define the scope of a search for information from a tabular data source
US8312553B2 (en) 2002-09-18 2012-11-13 Symantec Corporation Mechanism to search information content for preselected data
US8595849B2 (en) 2002-09-18 2013-11-26 Symantec Corporation Method and apparatus to report policy violations in messages
US9515998B2 (en) 2002-09-18 2016-12-06 Symantec Corporation Secure and scalable detection of preselected data embedded in electronically transmitted messages
US8813176B2 (en) 2002-09-18 2014-08-19 Symantec Corporation Method and apparatus for creating an information security policy based on a pre-configured template
US8751506B2 (en) 2003-05-06 2014-06-10 Symantec Corporation Personal computing device-based mechanism to detect preselected data
US20050195840A1 (en) * 2004-03-02 2005-09-08 Steven Krapp Method and system for preventing denial of service attacks in a network
US20070240207A1 (en) * 2004-04-20 2007-10-11 Ecole Polytechnique Federale De Lausanne (Epfl) Method of Detecting Anomalous Behaviour in a Computer Network
US8631464B2 (en) * 2004-04-20 2014-01-14 Ecole polytechnique fédérale de Lausanne (EPFL) Method of detecting anomalous behaviour in a computer network
US20060048019A1 (en) * 2004-08-26 2006-03-02 Canon Kabushiki Kaisha Image input/output device, information processing method, storage medium stored with program readable by computer, and program
US7653839B2 (en) * 2004-08-26 2010-01-26 Canon Kabushiki Kaisha Image input/output device, information processing method, and storage medium
US7607051B2 (en) * 2005-03-31 2009-10-20 Oki Electric Industry Co., Ltd. Device and method for program correction by kernel-level hardware monitoring and correlating hardware trouble to a user program correction
US7523359B2 (en) * 2005-03-31 2009-04-21 International Business Machines Corporation Apparatus, system, and method for facilitating monitoring and responding to error events
US20060248531A1 (en) * 2005-03-31 2006-11-02 Oki Electric Industry Co., Ltd. Information processing device, information processing method and computer-readable medium having information processing program
US20060230306A1 (en) * 2005-03-31 2006-10-12 Richards Paul M Apparatus, system, and method for facilitating monitoring and responding to error events
US20070055915A1 (en) * 2005-09-07 2007-03-08 Kobylinski Krzysztof R Failure recognition, notification, and prevention for learning and self-healing capabilities in a monitored system
US7702966B2 (en) * 2005-09-07 2010-04-20 Intel Corporation Method and apparatus for managing software errors in a computer system
US7823029B2 (en) * 2005-09-07 2010-10-26 International Business Machines Corporation Failure recognition, notification, and prevention for learning and self-healing capabilities in a monitored system
US20070055914A1 (en) * 2005-09-07 2007-03-08 Intel Corporation Method and apparatus for managing software errors in a computer system
US20070177523A1 (en) * 2006-01-31 2007-08-02 Intec Netcore, Inc. System and method for network monitoring
US20070245313A1 (en) * 2006-04-14 2007-10-18 Microsoft Corporation Failure tagging
US7673178B2 (en) 2007-01-31 2010-03-02 Microsoft Corporation Break and optional hold on failure
US7788540B2 (en) 2007-01-31 2010-08-31 Microsoft Corporation Tracking down elusive intermittent failures
US20080184079A1 (en) * 2007-01-31 2008-07-31 Microsoft Corporation Tracking down elusive intermittent failures
US20080184075A1 (en) * 2007-01-31 2008-07-31 Microsoft Corporation Break and optional hold on failure
US20080256149A1 (en) * 2007-04-13 2008-10-16 Nikhil Bansal System and method for dependent failure-aware allocation of distributed data-processing systems
US8122281B2 (en) * 2007-04-13 2012-02-21 International Business Machines Corporation System and method for dependent failure-aware allocation of distributed data-processing systems
US9235629B1 (en) 2008-03-28 2016-01-12 Symantec Corporation Method and apparatus for automatically correlating related incidents of policy violations
US8826443B1 (en) 2008-09-18 2014-09-02 Symantec Corporation Selective removal of protected content from web requests sent to an interactive website
US9118720B1 (en) 2008-09-18 2015-08-25 Symantec Corporation Selective removal of protected content from web requests sent to an interactive website
US20100152894A1 (en) * 2008-12-15 2010-06-17 Samsung Electronics Co., Ltd. Robot and recovery method thereof
US8892250B2 (en) * 2008-12-15 2014-11-18 Samsung Electronics Co., Ltd. Robot and recovery method thereof
US8613040B2 (en) * 2008-12-22 2013-12-17 Symantec Corporation Adaptive data loss prevention policies
US20100162347A1 (en) * 2008-12-22 2010-06-24 Ian Barile Adaptive data loss prevention policies
US8935752B1 (en) 2009-03-23 2015-01-13 Symantec Corporation System and method for identity consolidation
US8495422B2 (en) 2010-02-12 2013-07-23 Research In Motion Limited Method and system for resetting a subsystem of a communication device
US20110202797A1 (en) * 2010-02-12 2011-08-18 Evgeny Mezhibovsky Method and system for resetting a subsystem of a communication device
EP2642390A1 (en) * 2012-03-20 2013-09-25 BlackBerry Limited Fault recovery
US9026842B2 (en) 2012-03-20 2015-05-05 Blackberry Limited Selective fault recovery of subsystems

Also Published As

Publication number Publication date Type
US7386764B2 (en) 2008-06-10 grant
US20060143492A1 (en) 2006-06-29 application

Similar Documents

Publication Publication Date Title
Castelli et al. Proactive management of software aging
US6363495B1 (en) Method and apparatus for partition resolution in clustered computer systems
US5748882A (en) Apparatus and method for fault-tolerant computing
US6854069B2 (en) Method and system for achieving high availability in a networked computer system
US5684807A (en) Adaptive distributed system and method for fault tolerance
US7421478B1 (en) Method and apparatus for exchanging heartbeat messages and configuration information between nodes operating in a master-slave configuration
US6983324B1 (en) Dynamic modification of cluster communication parameters in clustered computer system
US5664088A (en) Method for deadlock recovery using consistent global checkpoints
US7730364B2 (en) Systems and methods for predictive failure management
Sullivan et al. Software defects and their impact on system availability-a study of field failures in operating systems
US6560720B1 (en) Error injection apparatus and method
US7024580B2 (en) Markov model of availability for clustered systems
US6061810A (en) Computer system with error handling before reset
US7321992B1 (en) Reducing application downtime in a cluster using user-defined rules for proactive failover
US7877642B2 (en) Automatic software fault diagnosis by exploiting application signatures
US6754855B1 (en) Automated recovery of computer appliances
US6789114B1 (en) Methods and apparatus for managing middleware service in a distributed system
US6901537B2 (en) Method and apparatus for preventing the propagation of input/output errors in a logical partitioned data processing system
US6952766B2 (en) Automated node restart in clustered computer system
US20040153724A1 (en) Operating system update and boot failure recovery
US7237140B2 (en) Fault tolerant multi-node computing system for parallel-running a program under different environments
US6505298B1 (en) System using an OS inaccessible interrupt handler to reset the OS when a device driver failed to set a register bit indicating OS hang condition
US20110131448A1 (en) Performing a workflow having a set of dependancy-related predefined activities on a plurality of task servers
Afek et al. Local stabilizer
US20030167421A1 (en) Automatic failure detection and recovery of applications

Legal Events

Date Code Title Description
AS Assignment

Owner name: SARVEGA, INC., ILLINOIS

Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNORS:LEDUC, DOUGLAS E.;LASH, JOHN K.;KOLLURU, NAGENDRA V.;REEL/FRAME:013082/0926

Effective date: 20020530

AS Assignment

Owner name: INTEL CORPORATION, CALIFORNIA

Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNOR:SARVEGA, INC.;REEL/FRAME:018163/0474

Effective date: 20060125

FPAY Fee payment

Year of fee payment: 4

FPAY Fee payment

Year of fee payment: 8

FEPP

Free format text: MAINTENANCE FEE REMINDER MAILED (ORIGINAL EVENT CODE: REM.)

LAPS Lapse for failure to pay maintenance fees

Free format text: PATENT EXPIRED FOR FAILURE TO PAY MAINTENANCE FEES (ORIGINAL EVENT CODE: EXP.)

FP Expired due to failure to pay maintenance fee

Effective date: 20180214